Impact of HLA High-Resolution Matching on Outcomes of Myeloablative Unrelated Donor Hematopoietic Stem Cell Transplantation in Chinese Population

Unrelated donor hematopoietic stem cell transplantation (URD-HSCT) is more frequently associated with severe graft-versus-host disease (GVHD) or graft rejection, and the success of URD-HSCT is influenced by the degree of HLA compatibility between the donor and patient. However, HLA mismatched unrelated donors should to be considerable for patients awaiting allogeneic HSCT who lack a suitable related donor or matched unrelated donor. The purpose of the study was to observed the impact of HLA-A, -B, -DRB1/B3 high-resolution matching on outcomes of URD-HSCT in Chinese population.

Patients and methods: 182 patients with hematological malignancies received URD-HSCT (bone marrow, n=130; peripheral blood stem cell, n=52) in our center between Nov. 1998 and May. 2008, and donors were from Chinese Marrow Donor Program (Chinese Mainland) and Tzu Chi Stem Cells Center (Chinese Taiwan), and the median age of all patients was 26 years (range 8–52 years). The selection of unrelated donor relied on donor-recipient HLA-A, -B, -DRB1/B3 matching by high-resolution molecular typing by PCR-SSP or PCR-SSO, with 121 cases of HLA 6/6 alleles matched, 51 cases of 1/6 allele mismatched and 10 cases of 2/6 alleles mismatched. The distribution of single HLA class I or class II mismatching was as follows: 37 HLA class I mismatching with 21 HLA-A and 16 HLA-B, and 14 HLA class II mismatching with 12 HLA-DRB1 and 2 HLA-DRB3. All of the patients were received Bu/Cy or Bu/Cy modified myeloablative conditionging regimen. MMF combined with CsA and short course MTX were performed as aGVHD prophylaxis, while other 18 patients received additional anti-CD25 monoclonal antibody to prevent severe aGVHD.

Results: After a median follow-up of 14.9 months, 170 patients achieved sustained engraftment with the engraft failure of 6.6%, early treatment-related mortality (TRM) of all patients was 14.4% at 100 days after transplant, and clinical relapse was observed in 8 patients (16.5%). aGVHD developed in 106 (58.2%) patients of all with grade I–II 82 (45.1%) and grade III–IV 24 (13.1%). By Kaplan-Meier method, the accumulative probability of 5-year overall survival (OS) and disease free survival (DFS) of all patients was 51.65±4.15% and 47.38±4.05%, respectively. The incidences of aGVHD was a little higher in HLA 1–2 alleles mismatched group (n=61) compared to HLA matched group (n=121) (67.2% vs 53.7%, p>0.05), and the incidences of grades I–II and III–IV aGVHD in HLA mismatched transplants were 45.9% and 21.3% respectively, while those in HLA matched transplants were 44.6% and 9.1% respectively. Comparing the outcomes between HLA 1–2 alleles mismatched and HLA matched transplants, the engraft failure were 9.8% and 5.0% (P>0.05), and early TRM were 18.0% and 12.4% (P>0.05), respectively. The Kaplan-Meier probability OS at 5 years were 44.31±6.86% and 55.66±5.11% in HLA mismatched and matched group (P>0.05) respectively. In HLA 2 alleles mismatched URD-HSCT, the incidence of engraft failure and aGVHD were 30.0% and 80.0%, and the outcomes were really inferior to HLA matched transplants. The impact of single HLA class I (n=37) or HLA class II mismatched (n=14) on the results of URD-HSCT had been also studied, and incidences of aGVHD in HLA class I or class II mismatched transplants was not significantly different compared with HLA matched transplants. In HLA class I and class II mismatched URD-HSCT, the engraft failure were 5.4% and 7.1% (p>0.05), and early TRM were 13.5% and 35.7% (p>0.05), respectively. The probability OS at 5 years in single HLA class II mismatched transplants was significantly lower compared with HLA matched transplants (23.81±12.94% vs 55.66±5.11%, p<0.01).

Conclusion: URD-HSCT could be optimized by comprehensive and precise donor-recipient alleles matching, however, HLA mismatching was associated with the risk of URD-HSCT. Moreover, HLA 2 alleles mismatches of donor-recipient HLA-A, B, DRB high-resolution matching was correlated with an inferior clinical outcome. For patients with high-risk diseases without a suitable matched unrelated donor, alternative methods to URD-HSCT with a single HLA mismatch may permit early treatment before disease progression. In our study, it also demonstrated that HLA class I mismatching was correlated with a high incidence of aGVHD, and HLA class II mismatching was associated with an inferior overall survival in Chinese population, however, larger studies would have to dissect out the magnitude of the risk incurred with specific mismatches more clearly owing to small patient numbers in each group.